Papers with memory-constrained devices

6 papers
FlexiGPT: Pruning and Extending Large Language Models with Low-Rank Weight Sharing (2025.naacl-long)

Copied to clipboard

Challenge: Empirical evaluations demonstrate substantial performance gains over existing methods .
Approach: They propose a method to prune LLMs that selectively prunes model blocks based on an importance score and replaces them with a low-parameter replacement strategy.
Outcome: The proposed method achieves state-of-the-art performance on 5/6 and 6/6 benchmarks with a compression rate of 30% and 40%.
Memory-Efficient Structured Backpropagation for On-Device LLM Fine-Tuning (2026.acl-industry)

Copied to clipboard

Challenge: Existing approaches for fine-tuning large language models require a trade-off between exact gradients with high memory and low memory with noisy estimates (MeZO).
Approach: They propose a method which derivates gradients from LoRA's low-rank structure and manually deriving backward passes to exploit the low-level structure.
Outcome: The proposed method reduces peak memory from 361MB to 136MB for Qwen2.5-0.5B, enabling fine-tuning scenarios previously infeasible on memory-constrained devices.
Revisiting Offline Compression: Going Beyond Factorization-based Methods for Transformer Language Models (2023.findings-eacl)

Copied to clipboard

Challenge: Recent transformer language models achieve outstanding results on many downstream tasks, but their enormous size often makes them impractical on memory-constrained devices.
Approach: They propose an offline compression approach that reduces the complexity of the model by enabling collaboration between modules.
Outcome: The proposed approach outperforms commonly used factorization-based offline compression methods on various NLP tasks.
SwapMoE: Serving Off-the-shelf MoE-based Large Language Models with Tunable Memory Budget (2024.acl-long)

Copied to clipboard

Challenge: Mixture of experts (MoE) is a popular technique to improve capacity of Large Language Models (LLMs) but memory-constrained devices are a major concern in edge AI training and serving.
Approach: They propose a framework for efficient serving of MoE-based large language models with tunable memory budgets.
Outcome: Experiments show that SwapMoE can reduce memory consumption while maintaining reasonable accuracy.
Adaptive Compression of Word Embeddings (2020.acl-main)

Copied to clipboard

Challenge: Distributed representations of words have been an indispensable component for natural language processing (NLP) tasks.
Approach: They propose a method that uses a code-book approach to represent words as discrete codes such as (8, 5, 2, 4).
Outcome: The proposed method makes the highly compressed word embeddings without hurting the task accuracy.
MulCode: A Multiplicative Multi-way Model for Compressing Neural Language Model (D19-1)

Copied to clipboard

Challenge: a large number of parameters dominate the memory usage of deep neural nets . compression of embedding layers is the key to reducing memory usage .
Approach: They propose a multi-way multiplicative neural compressor to compress embedding layers . they use an adaptively created matrix and multiplicativative compositions to learn them .
Outcome: a new multi-way multiplicative neural compressor can achieve 41.38 times compression rate with little loss in performance.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations